Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476651Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Ravie Chandren Muniyandi, Dr. | |
| dc.contributor.author | Esraa Saad Abdulmajed (P75892) | |
| dc.date.accessioned | 2023-10-06T09:23:17Z | - |
| dc.date.available | 2023-10-06T09:23:17Z | - |
| dc.date.issued | 2019-03-05 | |
| dc.identifier.other | ukmvital:123668 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476651 | - |
| dc.description | Intrusion detection (ID) is the process of identifying anomalies over networks. Intrusions could take different forms such as Denial-of-Service (DoS), probing, flooding and others. Several researchers have addressed the task of ID by using the traditional machine learning techniques. Recently, several researchers have tended to use modified versions of Artificial Neural Network such as Extreme Learning Machine (ELM) and Online Sequential Extreme Learning Machine (OSELM). The key characteristic behind such techniques lies on its ability to provide a generalized topology for the hidden-output weights. This has contributed toward reducing the training time. Yet, both algorithms still suffer from the randomization among the input-hidden weights. Therefore, this study aims to propose a Backpropagation Online Sequential Extreme Learning Machine (BOSELM). The proposed method attempts to tune the input-hidden weights in order to improve the intrusion detection accuracy. Two datasets were used in this study including the KDD Cup 99 and its improved version the NSL-KDD 2009. The proposed method has been compared with both Support Vector Machine, ELM and OSELM classifiers. Experimental results showed that the proposed method has outperformed the other classifiers in terms of classification accuracy by achieving an f-measure of 99.74% for the first dataset and 75.57% for the second dataset. Although the proposed method showed a training time that is longer than both ELM and OSELM however, it has a relatively similar performance. This emphasizes the advantage of using BOSELM in terms of improving both efficiency and effectiveness in the task of intrusion detection.,Master of Information Technology | |
| dc.language.iso | eng | |
| dc.publisher | UKM, Bangi | |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
| dc.rights | UKM | |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | |
| dc.subject | Dissertations, Academic -- Malaysia | |
| dc.subject | Intrusion detection | |
| dc.subject | Machine learning | |
| dc.title | Backpropagation online sequential extreme learning machine for intrusion detection | |
| dc.type | theses | |
| dc.format.pages | 110 | |
| dc.identifier.barcode | 005800(2021)(PL2) | |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_123668+SOURCE1+SOURCE1.0.PDF Restricted Access | 1.61 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.